ExDocApi/ Docs

Introduction

ExDocApi turns any document into structured JSON. Upload a file, describe what you want, get clean data back — no templates, no training, no configuration.

What it does

You define the fields. The API extracts them.

POST /api/v1/extract
  file    → your document (PDF or image)
  extract → { fields: { key: "what to look for" } }

The result is a JSON object with exactly the keys you asked for:

{
  "invoice_number": "INV-2024-0042",
  "vendor_name":    "Acme Supplies Ltd.",
  "total":          "18,450.00",
  "due_date":       "2024-12-01"
}

How the pipeline works

Every extraction goes through a two-stage pipeline:

  Your File (PDF or image)
       |
       v
  [ Stage 1: Document Parser ]
    Converts to structured text
    Handles tables, layouts, scans
       |
       v
  [ Stage 2: Extraction Engine ]
    Finds your fields in the text
    Returns structured JSON
       |
       v
  Your Result { field: value }

Processing is asynchronous. You submit a job, get a job_id, then poll until done.

Job lifecycle

  Submit
    |
    v
  queued --> pending --> processing --> done
                                  |
                                  +--> failed

| Status | Meaning | |--------|---------| | queued | Waiting in queue to be picked up | | pending | Assigned to a worker, document parsing in progress | | processing | Fields being extracted from parsed document | | done | Complete — result is ready | | failed | Something went wrong — check the error field |

Key capabilities

Any document type. Invoice, medical report, contract, bank statement, passport, payslip — the same API handles all of them.

You define the schema. No preset templates to configure. Pass any field name with a plain-English description.

Flexible fields. Ask for 1 field or 50. Fields not found in the document return null, not an error.

Reusable schemas. Save field sets as named schemas and reference them across jobs.

Quick links

Next →Quickstart